Marine gravity is traditionally derived from altimeter data using physics-based methods, which are sensitive to altimeter data noise. Although deep learning methods can capture nonlinear relationships between marine gravity and multi-source data, purely data-driven approaches depend strongly on the quality and spatial distribution of training ship-borne gravity data. To combine the advantages of both methods, this study proposes a hybrid marine gravity recovery method that integrates the Inverse Vening Meinesz (IVM) method into a convolutional neural network (CNN) framework. Specifically, the IVM-derived marine gravity is incorporated into the CNN loss function as a constraint, and distance- and topography-dependent hyperparameters are introduced to control its contribution during training. The gravity model derived by the proposed IVM-constrained CNN method is validated using independent ship-borne gravity observations and a recognized marine gravity model. Assessed by ship-borne data, the proposed method improves accuracy by 28.5% and 6.8% compared with the IVM-derived and CNN-derived gravity models, respectively. Comparisons with the recognized model also indicate that the gravity model derived by the proposed method achieves the highest accuracy among the three models. These results show that embedding IVM constraints into deep learning frameworks can significantly improve the accuracy and robustness of marine gravity recovery.

Significant wave height (SWH) is vital for maritime safety and oceanographic research. This study systematically evaluated the accuracy of the multi-mission along-track SWH product distributed by Copernicus Marine Service (CMEMS) with a focus on China’s coastal waters. Validation used 2022–2024 in situ data from 30 coastal operational buoys. A spatiotemporal collocation of 25 km and ±30 min was adopted to balance sample size and matching accuracy. The SWH product from CMEMS showed good performance with an overall bias of 0.10 m, root-mean-square error (RMSE) of 0.19 m, correlation coefficient (CC) of 0.98, and scatter index (SI) of 0.13. Spatially, errors tended to be lower in the south, offshore, and in deeper waters. All satellites systematically overestimated SWH under low sea states (SWH < 0.5 m) with a mean bias of 0.17 m, which was identified as a major contributor to the overall error. This study provides a practical reference for the validation of the multi-mission SWH product from CMEMS in China’s coastal waters, and also offers a useful reference for the fusion and assimilation of multi-source satellite wave data in complex coastal environments.

Postseismic deformation observed by terrestrial Global Navigation Satellite System (GNSS) and seafloor GNSS-Acoustic techniques (GNSS-A) provides essential constraints on the depth-dependent viscoelastic structure of subduction zones. In this study, we collect and process decadal postseismic observations following the 2011 Tohoku-oki M_w9.0 earthquake, including 232 onshore GNSS stations and six offshore GNSS-A sites. After removing the interseismic velocity terms, we extract the postseismic deformation signals mainly driven by viscoelastic relaxation during the period from 3 to 9 years after the earthquake. The inversion is primarily constrained by horizontal displacements, which have higher accuracy than vertical observations. We adopt a radially layered viscoelastic Earth model with lateral heterogeneity between continental and oceanic domains based on the Burgers rheology and half-space dislocation theory. Using the least-squares principle, we invert for the optimal viscoelastic structure under the strong constraint of fixed mantle viscosity. The optimal continental and oceanic crustal elastic thicknesses are 24.4 km and 37 km, with minimum horizontal Root-Mean-Square errors (RMS) of 5.68 cm and 6.81 cm, respectively. The mantle viscosity shows significant depth-dependence and obvious land–ocean differences. These results verify the critical role of joint land and seafloor geodetic constraints and provide a refined viscoelastic structure model for subduction zones.

Research on the seaward shoreface boundary (SSB) over broad spatial scales is limited, restricting the understanding of its macroscale characteristics. In this study, we define a theoretical wave-base-derived seaward shoreface boundary (TWSSB) based on the intersection of wave base (half the wavelength) with seabed topography. Using high-resolution wave analysis data from the Copernicus Marine Environment Monitoring Service (CMEMS) for 2022-2023, annual and seasonal wave bases were calculated in China’s coastal waters. By combining these data with nearshore bathymetry, the annual and seasonal TWSSBs were determined. The spatial distribution of annual TWSSBs and the migration of seasonal TWSSBs along China’s coast were detailed, providing a comprehensive view of TWSSB characteristics. Results show a strong correlation between TWSSBs and local isobaths. The annual TWSSBs approximately coincide with the 10 m isobath in the Bohai Sea and Beibu Gulf, align with the 20 m isobath in the central and southern Yellow Sea, and are close to the 30 m isobath east of Hainan Island. Compared to the annual averages, seasonal TWSSBs show different horizontal shifts, typically within several kilometers (median ~3 km). The maximum seasonal range (difference between seaward and landward shifts) reaches approximately 41 km in the South China Sea. Together, these findings present a large-scale synoptic mapping of the TWSSB along the entire coast of China and demonstrate that TWSSB location is influenced by regional bathymetry and seasonal wave conditions. This theoretical, wave-based approach is expected to offer a reference for future studies on SSB and coastal morphodynamic processes.